[OneDev #4] feat(tooling): refine X (Twitter) bookmarks scanning + scraping #28

Closed
opened 2026-05-24 04:46:39 +00:00 by joseph · 1 comment
Owner

Imported from OneDev issue #4 (id 35180)
Original project: internal-joseph
Original state: Closed
Original submitterId: 5
Original submitDate: 2026-04-08T14:37:18.497+00:00
Original lastActivity: {'date': '2026-04-18T12:22:44.866+00:00', 'description': "changed state to 'Closed'", 'userId': 5}

Original fields:

  • Type: New Feature
  • Priority: Normal
  • Assignees: None

Idea

Improve the X bookmarks ingestion pipeline — make it more reliable, more useful, and more queryable.

Likely improvements (refine after a current-state pass)

  • Auth flow stability (X breaks scrapers regularly — expect cookie/session refresh logic)
  • Incremental scraping (only new bookmarks since last run, not full re-pull)
  • Normalize into a structured store (SQLite? JSONL?)
  • Tag / topic extraction via local LLM (Lemonade Coder-30B)
  • Search interface — CLI or web UI tied into existing dashboards
  • Backup/export so a banned account doesn't lose history

Next session handoff notes

  • TBD by Joseph: where the current scraper lives (repo + last-run state)
  • TBD by Joseph: which of the improvements above are highest leverage right now
  • Likely uses Lightpanda or Playwright (both already deployed on VMID 128)

Why

Bookmarks are a high-signal personal knowledge stream that's currently locked inside X's UI. Every refinement compounds.

Imported from OneDev issue #4 (id 35180) Original project: internal-joseph Original state: Closed Original submitterId: 5 Original submitDate: 2026-04-08T14:37:18.497+00:00 Original lastActivity: {'date': '2026-04-18T12:22:44.866+00:00', 'description': "changed state to 'Closed'", 'userId': 5} Original fields: - Type: New Feature - Priority: Normal - Assignees: None ## Idea Improve the X bookmarks ingestion pipeline — make it more reliable, more useful, and more queryable. ## Likely improvements (refine after a current-state pass) - Auth flow stability (X breaks scrapers regularly — expect cookie/session refresh logic) - Incremental scraping (only new bookmarks since last run, not full re-pull) - Normalize into a structured store (SQLite? JSONL?) - Tag / topic extraction via local LLM (Lemonade Coder-30B) - Search interface — CLI or web UI tied into existing dashboards - Backup/export so a banned account doesn't lose history ## Next session handoff notes - **TBD by Joseph**: where the current scraper lives (repo + last-run state) - **TBD by Joseph**: which of the improvements above are highest leverage *right now* - Likely uses Lightpanda or Playwright (both already deployed on VMID 128) ## Why Bookmarks are a high-signal personal knowledge stream that's currently locked inside X's UI. Every refinement compounds.
Author
Owner

Imported from OneDev comment id 372 on issue #4
Original userId: 5
Original date: 2026-04-18T12:22:43.866+00:00

Closing as out of scope for this tracker — X bookmarks is a personal project, not PCT infrastructure. Can be reopened in a personal workspace if pursued.

Imported from OneDev comment id 372 on issue #4 Original userId: 5 Original date: 2026-04-18T12:22:43.866+00:00 Closing as out of scope for this tracker — X bookmarks is a personal project, not PCT infrastructure. Can be reopened in a personal workspace if pursued.
Sign in to join this conversation.
No labels
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
joseph/internal-joseph#28
No description provided.